Epidemiology of the Cerebral

Palsies
3
Eve Blair, Christine Cans, and Elodier Sellier

Abstract
Epidemiology of CP aims to describe the frequency of the condition in
a population and to monitor its changes over time. Also it studies the
determinants of this condition which responsible for some changes over
time. Classification of CP is an important step toward describing more
homogenous subgroups of persons with CP. Several classifications exist
based on neurological signs and topography, on motor function loss, on
associated impairments, on severity of the clinical pattern and on the neu-
roimaging findings. Overall prevalence of CP is around 2 per 1000 live
births in developed and in developing countries, with a trend toward a
decrease during the last decade, at least for the more severe subgroups
and the more tiny babies. Caution should be paid when interpreting
changes in prevalence rates since factors that may influence these esti-
mates are numerous.

Cerebral palsy (CP) is an umbrella term for a

E. Blair (*)
Telethon Kids Institute, University of Western
Australia, Perth, WA, Australia
e-mail: Eve.Blair@telethonkids.org.au
C. Cans
Universite Joseph Fourier Grenoble,
Grenoble, France
E. Sellier
Département de l’Information Médicale,
Pôle Santé Publique, Pavillon Taillefer,
Grenoble Cedex 9, France
­collection of conditions resulting in lifelong motor
disability. Together they constitute the most com-
mon cause of physical impairment in children and
are responsible for permanent ­activity l­imitation
and lifetime participation restriction [1]. CP is
defined by motor impairment due to maldevel-
opments of, or injury to, the immature brain,
the latter being a phrase open to variable inter-
pretation but often taken to be before 2 postnatal
years; see Chap. 2. Children and adults included
under the CP umbrella have very heterogeneous
clinical presentations. In addition to motor
impairments, they may also have impairments of

© Springer International Publishing AG 2018 19

C.P. Panteliadis (ed.), Cerebral Palsy, https://doi.org/10.1007/978-3-319-67858-0_3
20
cognition, vision, hearing, communication, pro-
prioception, behaviour and/or epilepsy that may
exacerbate or even overshadow the disability
due to the motor impairment. By definition the
responsible cerebral pathology is not progressive,
but the clinical manifestations change over time
with d­ evelopment and the ­long-term ­deleterious
effects of motor impairment on the musculoskel-
etal system. Those with severe CP tend to have
significantly lower survival rates, especially if
oromotor impairment is present.
Epidemiology is the study of the distribution
and of the determinants of disease frequency in
a population. The aims of epidemiology of CP
are multiple. Firstly it aims to evaluate the bur-
den of a disease by describing its frequency
according to its clinical characteristics. Secondly
it monitors trends in prevalence over time in the
search for clues regarding aetiology and preven-
tion. These trends are considered an important
measure of the impact of developments in pre-,
peri- and neonatal care since more than two
thirds of CP cases are considered to have aeti-
ologies involving these periods [2, 3]. Thirdly it
seeks specific risk factors and combinations of
risk factors associated with CP or different sub-
types of CP.
Following the first step of defining and describ-
ing CP (see Chap. 2), two other major steps are
(1) to collect information in order to be able to
study the frequency of the disease and (2) to seek
explanations for any changes in frequency.
3.1  efinitions About Rate
D the Ascertainment
and Trends, Methodological
Issues
Prevalence refers to the proportion of a popula-
tion with a condition and is most relevant for ser-
vice provision. It is calculated as the number of
persons with a condition (the numerator) divided
by the number of persons in the population from
which the numerator arose (the denominator).
Since both numerator and denominator change
over time, prevalence must be estimated at one
point in time (referred to as point prevalence)
which may or may not be stable over time. A
E. Blair et al.
trend in prevalence refers to a systematic change
in prevalence over a period of time.
Incidence refers to the number of new cases of
a condition arising per unit time. However since
this is of little epidemiological interest (unless
the size of the source population is known),
­incidence is often used as a shorthand for inci-
dence density rate which is the ratio of new cases
of a condition to the number of person-time units
at risk, where time is usually measured in years.
Incidence (density rate) is therefore more rele-
vant to aetiological investigation.
The relationship between prevalence and inci-
dence is simplified for a lifelong condition, since
an individual contributes only once to incidence
and will contribute to prevalence for the duration of
their survival. The problem with a condition such
as CP, which may not be recognised at the point
of acquisition, is that death may precede recogni-
tion. This is of little interest to service providers
who routinely divide the number of persons surviv-
ing with the CP label, typically to 2 years of age,
by the number of births from which they arose, a
ratio sometimes termed “birth prevalence”. This
will not estimate true prevalence if any losses or
gains occur in the source population before the
age at which the numerator is ascertained (due to
death or migration) but is nonetheless the method
used by the majority of CP surveillance systems [4]
which typically exist in developed countries where
neonatal and infant mortality is low.
3.2 Factors Affecting
Cerebral palsy registers have been established
around the world with the aim of estimating preva-
lence rates of CP in a population [5]. The first CP
registers set up in Europe were in Sweden (1954),
England (1966), Ireland (1966) and Denmark
(1967). Then, other registries were created in other
European countries. In 1999, a network of regis-
ters was established—gathering 14 registers or
population-based surveys from 8 different coun-
tries [6]. After harmonisation of their data, these
registers set up a common database. At present,
the network includes 24 active registers. In Western
3  Epidemiology of the Cerebral Palsies 21

Australia, a CP register was created in 1975, fol-
lowed by registers in the States of South Australia
and Victoria. Recently, new registers were imple-
mented in other parts of Australia, and a common
database, the Australian Cerebral Palsy Register
(ACPR), has been ­created, gathering data from all
registers [7] and assisting in the creation of com-
patible registers in New Zealand and Bangladesh.
In Japan, a CP register has been collecting data for
children born since 1988 in the Okinawa Prefecture
[8]. In the USA, the Autism and Developmental
Disabilities Monitoring Network (ADDM) [9]
includes three sites where population-based sur-
veys of CP are periodically held. Several excellent
surveys have been conducted in other parts of the
world including house-to-house data collection in
developing countries, demonstrating that the child-
hood prevalence of CP does not differ greatly
throughout the world; see Table 3.1. However
detailed examination of these surveys is beyond
the scope of this chapter.
Registers are population databases utilising
information from multiple sources, relying on a
clear definition and inclusion and exclusion crite-
ria of CP [5]. The use of multiple sources seeks
to maximise the accuracy and completeness of
data collection in a defined area (county, region or
country) and minimise selection bias that may arise
from hospital- or clinic-based studies. Population-
based studies also use multiple sources but dif-
fer from registers in that the inclusion of cases is
not continuous over time but limited usually to a
predefined span of birth dates. Although indepen-
dently created registers vary in methodology, simi-
larities exist regarding their aims, definitions and
the data they collect ­creating the possibility for

Table 3.1  Most recent prevalence estimates of cerebral palsy in different geographical locations
Study Children Birth
Reference Location population with CP (n) cohort Denominators Prevalence rate
Sellier et al. 20 registers in At least 10,756 1980–2003 Per 1000 live 1.90 in 1980 to
[10] Europe 4 years old births 1.77 in 2003
ACPR [7] 3 registers, 5 years old 3662 1993–2009 Per 1000 live 2.0 in 1993–
Australia births 1994 to 1.6 in
2007–2009
Van Naarden Metropolitan 8 years old 766 1985–2002 Per 1000 1-year 1.9 in 1985 to
Braun et al. Atlanta, USA survivors 2.2 in 2002
[11]
Touyama et al. Japan >2 years old 639 1988–2007 Per 1000 live 1.8 in 1988–
[8] births 1997 and 2.0 in
1998–2007
Liu et al. [12] China, seven <7 years old 622 1990–1997 Per 1000 children 1.6
cities in Jiangsu <7 years old
living in
surveillance areas
Yam et al. [13] Hong Kong 6–12 years old 578 1991–1997 Per 1000 children 1.3
enrolled in the
education system
in the school year
2003/2004
Smith et al. British Columbia, At least 497 1991–1995 Per 1000 live 2.7
[14] Canada 3 years old births
Oskoui et al. Quebec, Canada 9–11 years old 228 1999–2001 Per 1000 children 1.8
[15, 16] living in
surveillance areas
Oztürk et al. Turkey 2–16 years old 102 1990–2004 Per 1000 live 1.1
[17] births
Banerjee et al. Metropolis of <19 years old 48 1984–2003 Sample of 16,979 2.8
[18] Kolkata, India children
<19 years old
22
valid data comparisons and research collaborations
[4]. Numerous factors may affect the probabil-
ity that a child will be identified by a CP register.
These factors include eligibility criteria, age of
ascertainment, consent requirements and methods
used to identify new cases. Nevertheless, despite
these limitations, CP registers constitute invaluable
sources of detailed information for monitoring CP
characteristics, rates and trends and for the plan-
ning of services. They are useful also in the endless
search for causal pathways to CP.
3.3  lassification of Cerebral
C One more simple classification system was
Palsy
Clinical presentation is very heterogeneous with
examples found at all points of several continua,
though clinical descriptions do tend to cluster. It
is these clusters that are responsible for the tra-
ditional classification systems by type of motor
impairment, bodily distribution and severity. For
epidemiological purposes it is convenient to define
subgroups of individuals with a significant degree
of clinical homogeneity, but it must be appreciated
that such divisions are somewhat artificial; clini-
cal heterogeneity remains within subgroups, and
examples exist that defy such classification.
3.3.1 Classification Based
on Neurological Signs
and Topography of Motor
Impairment
The different classification systems based on
clinical findings serve different functions, some
of them having been proposed a long time ago
[19]. The main differences between classifica-
tions in use are the number of the different CP
types described. CP types have been given names
such as diplegia, quadriplegia, double hemiplegia,
triplegia, dystonic, dyskinetic, ataxic and mixed
mainly spastic [20, 21]. All these CP types are
quite useful for a clinician when describing the
condition of one child; however, for c­ omparison
purposes (often required in epidemiology), it has
been shown that the distinction between these
E. Blair et al.
terms, mainly between diplegia and ­quadriplegia,
is not reliable enough [22, 23]. None of these cat-
egories have a standardised definition in terms
of motor impairment, and their interpretation
is further complicated by variable assumptions
sometimes being made concerning non-motor
impairments. For instance, some professionals
will classify a child with CP diplegia type if the
two lower limbs are “predominantly” involved,
while others will classify the same child as quad-
riplegia because all four limbs are involved, and
for yet others the choice between these two cat-
egories may be determined by cognitive ability.
proposed recently by the European SCPE net-
work, with agreement between partners on the
clinical findings required for each following CP
type: bilateral spastic, unilateral spastic, dyski-
netic and ataxic. In the SCPE database, half of
children with CP have a bilateral spastic type,
nearly a third a unilateral spastic type, and the
others have either a dyskinetic or an ataxic type.
When more than one type is present, i.e. spastic-
ity with ataxia and/or dyskinesia, the child should
be classified according to the dominant clinical
feature [24]. This classification system had sub-
stantial to excellent interrater reliability [25] and
is recommended for epidemiological studies.
3.3.2 C
 lassification Based on Motor
Function Loss
The impact of the brain lesion on the motor func-
tion is often assessed through the ability of the
child to walk. It has been shown that walking
ability is strongly associated with CP type [26].
However, walking ability might mean walking
outdoors in everyday life, being able to walk
just a few metres or able only to walk indoors.
During the last decades, scales have been devised
specifically for children with CP to assess the
loss of motor function in both the lower and the
upper limbs, with the aim of increasing the com-
parability of assessments between profession-
als and geographically diverse research groups.
A five-­point scale for the trunk and lower limb
motor function, i.e. the Gross Motor Function
3  Epidemiology of the Cerebral Palsies
Classification System (GMFCS) (see also chap-
ter “Integrated management with Botulinum
Neurotoxin A”), was proposed and validated by
the Canadian research group CanChild ([27],
www.canchild.ca). It was subsequently extended
and revised [28]. Similarly, two five-point scales
for fine motor abilities of the upper limbs have
been proposed: the Manual Ability Classification
System (MACS) and the Bimanual Fine Motor
Function (BFMF) scales [29, 30]. It is possible
to take into account the asymmetry of the upper
limb function with the BFMF scale which also
provides complementary information about the
fine motor capacity [31]. Within these broad
categories, human functional abilities are mul-
tidimensional, so it should come as no surprise
that the two scales that purport to measure upper
limb function differ somewhat in which aspects
of upper limb function are assessed, as do the
two scales assessing communication. Moreover,
both GMFCS and BFMF scales allow the level
to be determined from medical records, and par-
ents have been shown to be able to reliably assess
their child’s GMFCS level [32]. The use of these
scales represents a great step forward in describ-
ing a child with CP and has greatly increased the
validity of comparing results between registers
and surveys.
3.3.3 Classification Based
on Associated Impairments,
Including Epilepsy
As mentioned in the most recent CP definition
[33], a child with CP may present several other
associated neurosensorial impairments. Many
of them are well defined and quantified, and
detailed descriptions are given in Chap. 10 on
comorbidities. For some such as communica-
tion and speech disorders and feeding problems,
standardised assessment tools have only very
recently been proposed [34–36], and population-
based studies assessing their frequency among
children with CP are lacking. The use of these
standardised tools to collect information on all
these comorbidities is highly recommended both
for epidemiological purposes and to achieve
23
o­ ptimal outcomes in children with CP and their
families [37].
Intellectual impairment is the most com-
monly associated impairment encountered among
children with CP. It occurs in 30–40% when
restricted to severe intellectual impairment (IQ
level < 50) [38] and up to 60–70% if “mild” intel-
lectual impairment or any specific learning dis-
abilities are also included. On account of their
other impairments, it is not always easy to appre-
ciate the intellectual level in children with CP, and
it is probably quite often underestimated [39].
The second most frequently associated impair-
ment is epilepsy which may appear early in life
or not until school age. Many studies report that
about 30% of people with CP have active epi-
lepsy varying in type depending on the type of
the brain anomaly [40]. Some will require con-
tinuous long-term treatment in order to avoid
repeated fits compounding the original cerebral
anomaly.
The third most frequent associated impair-
ment is visual impairment which can be a squint
and/or a loss of visual acuity. Several studies
report that about 15% of children with CP have a
severe visual impairment (visual acuity below
0.01 on the better eye after correction) [41].
Severe hearing impairment, defined as a loss
of 70 db or more in the better ear before correc-
tion, is rare in children with CP at around 2–3%.
Children with CP may also present with
behavioural disorder, as a direct consequence of
the brain lesion or as a consequence of their activ-
ity limitation and participation restriction. Severe
psychiatric conditions such as autism are also
infrequently observed in children with CP [42].
3.3.4 Severity Assessment
in Children with CP
Using the classification systems described
above, it may be possible to say that one child
with CP presents with a more severe pattern of
any particular facet of their impairments than
another. This is very important when comparing
study samples whether considering prevalence
of CP, assessing medical or educational needs or
24
the appropriateness or efficacy of management
interventions.
Clinicians usually consider severity in terms
of loss of motor ability, possibly in combination
with type of motor disorder [43], and the wide-
spread use of GMFCS, BFMF and MACS have
greatly increased the comparability of results
between centres [44]. In Australia the Australian
Spasticity Assessment Scale is used to categorise
the severity of spastic impairment [45]. Another
scale has been developed for dyskinesia based
on the Barry-Albright Dystonia (BD) scale [46,
47]. In some studies motor function and intel-
lectual impairment levels are used together to
define mild, moderate and severe impairment
[48–50], and in a study of life expectancy, the
type and severity of motor impairment, together
with impairments of intellect, epilepsy, sight
and hearing, were combined into a 12-category
overall disability score [51]. Ideally, the assess-
ment of overall disability in the person with CP
would be assessed by combining the standardised
assessments of the many possible components of
impairment. The challenge is to know how to
weight the different components, since the contri-
bution to disability of each individual component
varies with the activities under consideration.
3.3.5 Classification Based
on Neuroimaging Results
Following recent recommendations, [52] cere-
bral MRI is performed more frequently on chil-
dren with CP, which together with improved
image quality has allowed classification sys-
tems of brain images to be proposed, e.g. [53,
54]. Systematic classification is of great interest
in epidemiological studies that seek to monitor
trends over time for different groups of children
with CP according to their cerebral pathology
and also provides important information about
the probable timing of the lesion responsible of
the CP [53]. Not all children with CP receive
cerebral MRI after the recommended 2 years of
age; thus, a standardised classification has also
been proposed for neonatal ultrasound imag-
ing results (https://pehta.sites.innovatif.com/fib.
E. Blair et al.
scpenet/en/my-scpe/rtm/neuroimaging/neona-
tal-neuroimaging/). In both MRI and neonatal
ultrasound imaging classifications, the main cat-
egories in the European system are maldevelop-
ments, predominant white matter injury (PVL,
IVH) and predominant grey matter injury (basal
ganglia, cortical subcortical, arterial infarc-
tions), while the Australian system separates
out the focal vascular insults and reports that
combinations of categories occurred rarely [54].
Population-based registers indicate a lesional
pattern in more than two thirds of MRI results,
more often predominantly in white than corti-
cal or deep grey matter. Maldevelopments were
observed in less than 10% of cases and nor-
mal findings in a bit more than 10% [53, 54].
Harmonisation of neuroimaging classifications
and international consensus are important for
the future [54], and training tools based on web
imaging will be very helpful for the physicians
and students [55]. Quantitative aspect may also
be taken into account for neuroimaging classifi-
cations, and several have been proposed recently
including one for unilateral spastic lesion [56]
and a semi-quantitative one for all children with
CP [57]; however these should currently be con-
sidered as research tools rather than tools for
clinical practice.
3.4  revalence of CP and Time
P
Trends
Prevalence estimates of CP are around 2 per 1000
live births (Table 3.1). Prevalence of CP is
strongly associated with gestational age and birth
weight. The rate is 1 per 1000 live births in chil-
dren born at term, 7–10 times higher for children
born moderately preterm (i.e. 32–36 weeks
­gestational age) and 60 times higher for children
born very preterm (i.e. before 32 weeks gesta-
tional age) [15, 16, 58]. Prevalence is higher in
males than female (rate ratio of 1.35) [59]. The
prevalence is also higher for children born in
multiple births, with risk increasing with increas-
ing multiplicity. Multiples tend to be born more
preterm than singletons, and their increased risk
of CP is mainly related to the higher risk of
3  Epidemiology of the Cerebral Palsies
p­ reterm birth in multiples even though there are
also causal factors specific to multiple pregnan-
cies such as cofetal demise and twin-twin trans-
fusion [60]. The prevalence varies according to
race/ethnicity with a higher prevalence rate for
non-­Hispanic black children than for non-His-
panic white children [11] and a higher rate for
indigenous than for nonindigenous children [61].
Population-­based studies on CP that have been
conducted in developing countries [12, 18] sur-
prisingly showed no large differences in preva-
lence rates between developed and developing
countries. Nevertheless, disparities in methodol-
ogy, age range, classification systems and popu-
lations make the studies difficult to compare to
those made in higher-income settings [62].
The prevalence of CP of postneonatal origin,
i.e. CP as a result of a recognised causal event
occurring between the 28th day and 24 months,
was estimated at 1.2 per 10,000 live births in
Europe (birth years 1976–1998) [63] with a
decrease over the period.
The oldest population-based registers have
monitored the prevalence rate of CP since the
1950s. During several decades, no systematic
time trend was observed apart from a slight
increase in the 1980s [64]. In more recent years,
SCPE has shown a decrease in CP prevalence for
children born between 1980 and 2003 in Europe
[10] with an important decrease for children born
with a birth weight between 1000 and 1499 g
(from 70.9 to 35.9/1000 live births), a significant
decrease for children with a birth weight between
1500 and 2499 g (from 8.5 to 6.2/1000 live
births) but no significant change for children born
with a birth weight below 1000 g or over 2500 g.
The decrease was observed primarily in bilateral
spastic CP and concerned all levels of severity. In
Australia, the ACPR reported, after a long period
of stable prevalence at around 2–2.5/1000 1000
live births, a decline to 1.4–2.1/1000 in the 2007–
2009 period [7]. Over the 2003–2009 period, the
rate of CP/1000 live births reduced across all ges-
tational age groups although there was a decrease
in the proportion with co-occurring intellectual
disability [11].
Moreover, studies on the different subtypes of
CP suggest different time trends according to
25
subtype. A decrease of bilateral spastic CP
[65,  66] and an increase of unilateral CP were
described for children born between 1983 and
1998 [66]. Prevalence of children with dyskinetic
CP has also risen between birth years 1976 and
1996 [48].
3.5  aution with Interpretation
C
of Trends
Factors that may influence CP prevalence rates
are numerous, and it is difficult to determine
which ones are related to the observed changes in
trends. Similarly, differences between estimated
rates and overtime are not easily understandable
even if we can suggest potential reasons. First,
numerous methodological issues may have an
impact on prevalence estimates including the type
of ascertainment sources (e.g. schools, registers,
administrative databases), the changes over time
or between countries in the inclusion or exclusion
criteria (e.g. inclusion of mild cases, postnatal
cases or children with isolated hypotonia) and the
denominators used. For a valid estimate of trend,
each prevalence estimate contributing to the
trend must be comparable. Even though registers
attempt to maintain constant methods of ascertain-
ment, changes in the external environment, such
as privacy legislation or restructuring the provi-
sion of services for children with CP, necessitate
“tweaking” ascertainment methods with the risk
of affecting proportional ascertainment. Second,
changes in the prevalence of determinants of CP
are almost certain to result in changes in CP prev-
alence. For example, decreases in stillbirth and
neonatal mortality rates are accompanied by an
increase of the number of surviving children born
extremely premature who remain at higher risk of
CP than children born at term despite declining
gestation-­specific CP rates.
Since CP includes conditions resulting from
many causal pathways together with the require-
ment for survival to an age at CP can be reliably
recognised, the constant prevalence observed over
many decades in developed countries and the
equivalence of prevalence between developed and
developing countries should not be interpreted as
26
constancy or equivalence of aetiological ­profile.
The period of static prevalence in developed
countries saw significant increases in the survival
of compromised births and a consequent rise in
their contributions to CP balanced by significant
decreases in CP resulting from now preventable
causes such as kernicterus secondary Rhesus
incompatibility and maternal rubella and postneo-
natally to motor vehicle accidents and cerebral
infections. In developing countries the aetiological
profile may more resemble that which existed pre-
viously in developed countries: neonatal mortality
removes compromised births, but the more techno-
logical preventive strategies may not be available,
and unselected home births occur more frequently.
While lower CP prevalence is seen as most
desirable from the human and service provision
points of view, that desirability may be question-
able if it comes at the cost of increased perinatal
mortality.
Conclusion
The concept of cerebral palsy evolved recently
with emphasis on the fact that it is very often
accompanied by other neurosensorial impair-
ments. Harmonisation in the way of describ-
ing CP led to improvements when monitoring
CP rates and when comparing trends between
countries and over time. However, several risk
factors may also change over time, including
those related to care delivered to pregnant
women and care to the newborns. Thus, sur-
veillance of CP should continue in both devel-
oped and developing countries.
References
1. World Health Organization. International classifi-
cation of functioning, disabilty and health (ICF).
Geneva: World Health Organization; 2001.
2. Himmelmann K, Hagberg G, Beckung E, et al. The
changing panorama of cerebral palsy in Sweden.
IX. Prevalence and origin in the birth-year period
1995–1998. Acta Paediatr. 2005;94:287–94.
3. Himmelmann K, Hagberg G, Uvebrant P. The
changing panorama of cerebral palsy in Sweden.
X. Prevalence and origin in the birth-year period
1999–2002. Acta Paediatr. 2010;99:1337–147.
E. Blair et al.
4. Goldsmith S, McIntyre S, Smithers-Sheedy H, et al.
On behalf of the Australian Cerebral Palsy Register
Group. An international survey of cerebral palsy
registers and surveillance systems. Dev Med Child
Neurol. 2016;58(Suppl 2):11–7.
5. Cans C, Surman G, McManus V, et al. Cerebral palsy
registries. Semin Pediatr Neurol. 2004;11:18–23.
6. Surveillance of Cerebral Palsy in Europe. Surveillance
of cerebral palsy in Europe: a collaboration of cerebral
palsy surveys and registers. Surveillance of Cerebral
Palsy in Europe (SCPE). Dev Med Child Neurol.
2000;42:816–24.
7. ACPR Group. Report of the Australian Cerebral Palsy
Register, birth years 1993–2009. 2016. www.cpregis-
ter.com
8. Touyama M, Touyama J, Toyokawa S, Kobayashi
Y. Trends in the prevalence of cerebral palsy in chil-
dren born between 1988 and 2007 in Okinawa, Japan.
Brain and Development. 2016;38:792–9.
9. Yeargin-Allsopp M, Van Naarden Braun K, Doernberg
NS, et al. Prevalence of cerebral palsy in 8-year-old
children in three areas of the United States in 2002: a
multi-site collaboration. Pediatrics. 2008;121:547–54.
10. Sellier E, Platt MJ, Andersen GL, et al. Decreasing
prevalence in cerebral palsy: a multi-site European
population-based study, 1980 to 2003. Dev Med Child
Neurol. 2016;58:85–92.
11. Van Naarden Braun K, Doernberg N, Schieve L, et al.
Birth prevalence of cerebral palsy: a population-based
study. Pediatrics. 2016;137:1–9.
12. Liu JM, Li S, Lin Q, Li Z. Prevalence of cerebral
palsy in China. Int J Epidemiol. 1999;28:949–54.
13. Yam WK, Chan HS, Tsui KW, et al. Prevalence study
of cerebral palsy in Hong Kong children. Hong Kong
Med J. 2006;12:180–4.
14. Smith L, Kelly KD, Prkachin G, Voaklander DC. The
prevalence of cerebral palsy in British Columbia,
1991–1995. Can J Neurol Sci. 2008;35:342–7.
15. Oskoui M, Coutinho F, Dykeman J, Jette N,
Pringsheim T. An update on the prevalence of c­ erebral
palsy: a systematic review and meta-analysis. Dev
Med Child Neurol. 2013a;55:509–19.
16.
Oskoui M, Joseph L, Dagenais L, Shevell
M. Prevalence of cerebral palsy in Quebec: alternative
approaches. Neuroepidemiology. 2013b;40:264–8.
17. Oztürk A, Demirci F, Yavuz T, et al. Antenatal and
delivery risk factors and prevalence of cerebral
palsy in Duzce (Turkey). Brain and Development.
2007;29:39–42.
18. Banerjee TK, Hazra A, Biswas A, et al. Neurological
disorders in children and adolescents. Indian J Pediatr.
2009;76:139–46.
19. Alberman E. Describing the cerebral palsies: methods
of classifying and counting. In: Stanley F, Alberman
E, editors. The epidemiology of the cerebral palsies,
Clinics in developmental medicine, vol. No. 87.
London: Spastics International Medical Publications;
1984. p. 27–31.
20. ACPR Group. Report of the Australian Cerebral Palsy
Register, birth years 1993–2003. 2009.
3  Epidemiology of the Cerebral Palsies
21. Stanley F, Blair E, Alberman E. Causal pathways to
the cerebral palsies: a new aetiological model. Chapter
5. In: Cerebral palsies: epidemiology and causal path-
ways. London: Mac Keith Press; 2000. p. 40–7.
22. Colver A. Benefits of a population register of chil-
dren with cerebral palsy. Indian Pediatr. 2003;40:
639–44.
23. Howard J, Soo B, Graham HK, et al. Cerebral palsy
in Victoria: motor types, topography and gross motor
function. J Paediatr Child Health. 2005;41:479–83.
24. Cans C, Dolk H, Platt MJ, et al. Recommendations
from the SCPE collaborative group for defining and
classifying cerebral palsy. Dev Med Child Neurol.
2007;109(Suppl):35–8.
25. Sellier E, Horber V, Krägeloh-Mann I, et al. Interrater
reliability study of cerebral palsy diagnosis, neuro-
logical subtype, and gross motor function. Dev Med
Child Neurol. 2012;54:815–21.
26. Beckung E, Hagberg G, Uldall P, Cans C. Probability
of walking in children with cerebral palsy in Europe.
Pediatrics. 2008;121:e187–92.
27. Palisano R, Rosenbaum P, Walter S, et al. Development
and reliability of a system to classify gross motor
function in children with cerebral palsy. Dev Med
Child Neurol. 1997;39:214–23.
28. Palisano RJ, Rosenbaum P, Bartlett D, Livingston
MH. Content validity of the expanded and revised
Gross Motor Function Classification System. Dev
Med Child Neurol. 2008;50:744–50.
29. Beckung E, Hagberg G. Neuroimpairments, activ-
ity limitations, and participation restrictions in chil-
dren with cerebral palsy. Dev Med Child Neurol.
2002;44:309–16.
30. Eliasson AC, Krumlinde-Sundholm L, Rösblad B,
Beckung E, Arner M, Ohrvall AM, Rosenbaum P. The
Manual Ability Classification System (MACS) for
children with cerebral palsy: scale development and
evidence of validity and reliability. Dev Med Child
Neurol. 2006;48:549–54.
31. Elvrum AK, Andersen GL, Himmelmann K, Beckung
E, Öhrvall AM, Lydersen S, Vik T. Bimanual fine
motor function (BFMF) classification in children with
cerebral palsy: aspects of construct and content valid-
ity. Phys Occup Ther Pediatr. 2016;36(1):1–16.
32. Morris C, Kurinczuk JJ, Fitzpatrick R, Rosenbaum
PL. Who best to make the assessment? Professionals
in cerebral palsy and families’ classifications of gross
motor function are highly consistent. Arch Dis Child.
2006;91:675–9.
33. Rosenbaum P, Paneth N, Leviton AB, et al. A report:
the definition and classification of cerebral palsy April
2006. Dev Med Child Neurol. 2007;109(Suppl):8–14.
34. Barty E, Caynes K, Johnston LM. Development
and reliability of the Functional Communication
Classification System for children with cerebral palsy.
Dev Med Child Neurol. 2016;58(10):1036–41.
35. Remijn L, Speyer R, Groen BE, et al. Validity
and reliability of the Mastication Observation and
Evaluation (MOE) instrument. Res Dev Disabil.
2014;35(7):1551–61.
27
36. Virella D, Pennington L, Andersen GL, Andrada Mda
G, Greitane A, Himmelmann K, Prasauskiene A,
Rackauskaite G, De La Cruz J, Colver A. Surveillance
of Cerebral Palsy in Europe Network. Classification
systems of communication for use in epidemiological
surveillance of children with cerebral palsy. Dev Med
Child Neurol. 2016;58:285–91.
37. Pruitt DW, Tsai T. Common medical comorbidities
associated with cerebral palsy. Phys Med Rehabil
Clin N Am. 2009;20:453–67.
38. Surveillance of cerebral palsy in Europe. Prevalence
and characteristics of children with cerebral palsy in
Europe. Dev Med Child Neurol. 2002;44:633–40.
39. Sigurdardottir S, Eiriksdottir A, Gunnarsdottir E,
et al. Cognitive profile in young Icelandic chil-
dren with cerebral palsy. Dev Med Child Neurol.
2008;50:357–62.
40. Carlsson M, Hagberg G, Olsson I. Clinical and aetio-
logical aspects of epilepsy in children with cerebral
palsy. Dev Med Child Neurol. 2003;45:371–6.
41. Venkateswaran S, Shevell MI. Comorbidities and
clinical determinants of outcome in children with
spastic quadriplegic cerebral palsy. Dev Med Child
Neurol. 2008;50:216–22.
42. Kilincaslan A, Mukaddes NM. Pervasive develop-
mental disorders in individuals with cerebral palsy.
Dev Med Child Neurol. 2009;51:289–94.
43. Shevell MI, Dagenais L, Hall N. The relationship of
cerebral palsy subtype and functional motor impair-
ment: a population-based study. Dev Med Child
Neurol. 2009;51:872–7.
44. Surman G, Hemming K, Platt MJ, et al. Children with
cerebral palsy: severity and trends over time. Paediatr
Perinat Epidemiol. 2009;23:513–21.
45. Love S, Gibson N, Smith N, et al. Interobserver
reliability of the Australian Spasticity Assessment
­
Scale (ASAS). Dev Med Child Neurol. 2016;58(Suppl
2):18–24.
46.
Barry MJ, VanSwearingen JM, Albright
AL. Reliability and responsiveness of the Barry-­
Albright Dystonia Scale. Dev Med Child Neurol.
1999;41(6):404–11.
47. Monbaliu E, Ortibus E, De Cat J, et al. The Dyskinesia
Impairment Scale: a new instrument to measure dys-
tonia and choreoathetosis in dyskinetic cerebral palsy.
Dev Med Child Neurol. 2012;54(3):278–83.
48. Himmelmann K, McManus V, Hagberg G, et al.
Dyskinetic cerebral palsy in Europe: trends in preva-
lence and severity. Arch Dis Child. 2009;94:921–6.
49. Platt MJ, Cans C, Johnson A, et al. Trends in cerebral
palsy among infants of very low birthweight (<1500
g) or born prematurely (<32 weeks) in 16 European
centres: a database study. Lancet. 2007;369:43–50.
50. Sellier E, Surman G, Himmelmann K, et al. Trends in
prevalence of cerebral palsy in children born ≥2500
g in Europe from 1980 to 1998. Eur J Epidemiol.
2010;25:635–42.
51. Blair E, Watson L, Badawi N, Stanley F. Life expec-
tancy among people with cerebral palsy in Western
Australia. Dev Med Child Neurol. 2001;43:508–15.
28
52. Ashwal S, Russman BS, Blasco PA, et al. Practice
parameter: diagnostic assessment of the child with
cerebral palsy: report of the Quality Standards
Subcommittee of the American Academy of
Neurology and the Practice Committee of the
Child Neurology Society. Neurology. 2004;62(6):
851–63.
53. Krägeloh-Mann I, Horber V. The role of magnetic
resonance imaging in furthering understanding of
the pathogenesis of cerebral palsy. Dev Med Child
Neurol. 2007;49(12):948.
54. Reid SM, Dagia CD, Ditchfield MR, et al. Population-­
based studies of brain imaging patterns in cerebral
palsy. Dev Med Child Neurol. 2014;56(3):222–32.
55. Platt MJ, Krageloh-Mann I, Cans C. Surveillance of
cerebral palsy in Europe: reference and training man-
ual. Med Educ. 2009;43(5):495–6.
56. Shiran S, Weinstein M, Sirota-Cohen C, Myers V,
Ben Bashat D, Fattal-Valevski A, Green D, Schertz
M. MRI-based radiologic scoring system for extent
of brain injury in children with hemiplegia. Am J
Neuroradiol. 2014;35:2388–96.
57. Fiori S, Cioni G, Klingels K, Ortibus E, Van Gestel
L, Rose S, Boyd R, Feys H, Guzetta A. Reliability
of a novel, semi-quantitative scale for classification
of structural brain magnetic resonance imaging in
children with cerebral palsy. Dev Med Child Neurol.
2014;56:839–45.
58. Cans C, de la Cruz J, Mermet MA. The epidemi-
ology of cerebral palsy. Paediatr Child Health.
2008;18:393–8.
59. Reid SM, Meehan E, Gibson C, Scott H, Delacy M,
On behalf of the Australian Cerebral Palsy Register
E. Blair et al.
Group. Biological sex and the risk of cerebral palsy
in Victoria, Australia. Dev Med Child Neurol.
2016;58:43–9.
60. Topp M, Huusom LD, Langhoff-Roos J, Delhumeau
C, Hutton JL, Dolk H, SCPE Collaborative Group.
Multiple birth and cerebral palsy in Europe: a
multicenter study. Acta Obstet Gynecol Scand.
2004;83:548–53.
61. Blair E, Watson L, O’Kearney E. Comparing risks of
cerebral palsy in births between Australian Indigenous
and non-Indigenous mothers. Dev Med Child Neurol.
2016;58(Suppl 2):36–42.
62. Gladstone M. A review of the incidence and preva-
lence, types and aetiology of childhood cerebral
palsy in resource-poor settings. Ann Trop Paediatr.
2010;30:181–96.
63. Germany L, Ehlinger V, Klapouszczak D, et al.
Trends in prevalence and characteristics of post-natal
cerebral palsy cases: a European registry-based study.
Res Dev Disabil. 2013;34:1669–77.
64. Paneth N, Hong T, Korzeniewski S. The descrip-
tive epidemiology of cerebral palsy. Clin Perinatol.
2006;33:251–67.
65. Himmelmann K, Beckung E, Hagberg G, Uvebrant
P. Bilateral spastic cerebral palsy - prevalence through
four decades, motor function and growth. Eur J
Paediatr Neurol. 2007;11:215–22.
66. Ravn SH, Flachs EM, Uldall P. Cerebral palsy in
eastern Denmark: declining birth prevalence but
increasing numbers of unilateral cerebral palsy in
birth year period 1986–1998. Eur J Paediatr Neurol.
2009;14:214–8.